The Detection of Donor-Directed, HLA-Specific Alloantibodies in Recipients of Unrelated Hematopoietic Cell Transplantation Is Predictive of Graft Failure.

Animal studies point to a strong role for MHC-specific antibody as a cause for failed donor hematopoietic cell engraftment, but the role of donor-directed HLA-specific allo-antibodies in human transplants has been controversial. To investigate such a role, we used a retrospective case-control design and studied unrelated donor:recipient pairs whose transplants were facilitated through the NMDP. A total of 37 cases with graft failure and 78 matched control pairs were evaluated. The 37 graft failure cases were selected based on survival beyond 28 days of transplant with no sustained engraftment, and available cryopreserved recipient serum to test for HLA-specific alloantibodies. Up to 3 controls were selected for each case, and matched for disease, disease status, graft type, patient age and year of transplant. Patients had AML, ALL, CML or MDS, 98% received myeloablative conditioning regimens, 100% received T replete grafts, 97% received marrow, and 97% received calcineurin-based GVHD prophylaxis. Patients and donors were retrospectively typed for HLA-A,B,Cw,DRB1,DQB1,DQA1,DPB1,and DPA1 by sequencing or other high resolution typing methods. Stored pre-transplant serum samples (patients and controls) were retrieved from the NMDP Research Repository, and assayed for HLA antibodies by solid-phase FlowPRA (One Lambda, Inc). All positive samples were evaluated for HLA specificity by single-antigen microparticles (LabScreen, One Lambda, Inc). Among the 37 failed transplants, 11 (30%) recipients possessed alloantibodies specific for donor HLA Class I or Class II, compared to only 3 (4%) of 78 controls. HLA-DR or DQ specific antibodies were not detected; hence, all HLA Class II-specific antibodies were directed to DP. Recipients with anti-DP antibodies against the donor mismatched DP were at increased risk of graft failure, indicating the importance of recipient antibodies directed against mismatched donor DP antigens. Exact conditional logistic regression analysis for the presence of either Class I HLA-A,B,Cw or Class II HLA-DP antibodies showed similar findings (Class I alone: OR. 6.31: 95% CI 1.17–62.9; p=0.03, Class II alone: OR 12.00; 95% CI 1.46–551.97; p=0.01, Class I and II combined: OR 19.08; 95% CI 2.72–828.49; p=0.0003). Further analyses were conducted to evaluate a limited set of covariates not accounted for in the case:control study, i.e. patient CMV status, cell dose and HLA-C match. Cell dose and CMV status were independently predictive of engraftment, p=0.01 and 0.03, respectively. No effect was observed for HLA-Cw match (p=0.84). The presence of anti-donor HLA Class I or II antibodies was predictive of engraftment when adjustment was made for either cell dose (OR 15.49; 95% CI 2.06–697.83; p=0.002) or CMV status (OR 7.94; 95% CI 0.97–367.84; p=0.05). In summary, these results indicate that donor-specific HLA Class I or Class II antibodies in recipients of unrelated donor hematopoietic cell transplants are associated with failed engraftment. We recommend that, as a “Standard-of-Practice”, all potential recipients be screened for the presence of HLA class I and class II antibodies including HLA DP. Donors should be excluded if they carry mismatched HLA types against which the patient has specific antibodies.